Clinical Utility

#Move from information to consequences

A test can produce accurate information without improving care. Clinical utility asks what happens when that information is used in a defined care pathway. The question concerns the whole sequence: who receives the test, how results influence decisions and what benefits or harms follow compared with a relevant alternative.

Useful effects might include better health outcomes, fewer unnecessary procedures or less burden while maintaining acceptable outcomes. Changes in clinician decisions can help explain how a test works, but decision changes alone do not prove patient benefit. The importance of each outcome should be made explicit.

#Compare realistic care strategies

The comparison should represent a credible alternative, such as the existing care pathway. Where feasible and ethical, randomized studies can help determine whether differences are caused by the test-guided strategy rather than other factors. Other designs may contribute evidence, but their susceptibility to confounding and selection needs careful assessment.

Studies should describe what actions follow each result and whether those actions occurred. Consider false alarms, missed conditions, unnecessary treatment, delays, distress and the burden of testing. Effects may depend on available treatments, staff capacity and access to follow-up, so utility is not a fixed property of a test in isolation.

#Distinguish observed effects from projections

Sometimes evidence links several steps rather than directly measuring final outcomes. For example, an evaluation may combine accuracy data with evidence about an established treatment. Such reasoning can be informative, but every link needs support. Uncertainty grows when assumptions concern who receives treatment or how well the pathway functions.

Decision models can explore possible benefits, harms and resource use, but their outputs depend on inputs and assumptions. Report projected effects separately from observed findings. A balanced utility assessment considers the size and certainty of benefits, unintended consequences and whether the results are likely to apply in the proposed setting.

#Common misunderstandings

A test can be accurate without being clinically useful. Correctly identifying a condition or predicting an outcome does not, by itself, show that using the result improves care. There must be a helpful action available, and the result must support better decisions than would otherwise be made.

More information is not automatically better. Additional findings can trigger unnecessary follow-up, anxiety, or treatment without improving outcomes that matter to patients. Equally, a reassuring result can cause harm if it delays appropriate assessment.

Clinical utility also does not mean that every patient benefits, or that a tool is useful in every setting. Benefits and harms can differ across patient groups, services, and stages of illness.

Finally, a change in clinicians’ decisions is not necessarily an improvement. Evidence that a tool changes referrals or prescribing needs to be connected to whether those changes are appropriate and produce worthwhile benefits after accounting for harms and burdens.

#Questions worth asking a clinician

• Compared with current care without this test or model, what evidence shows that using it improves outcomes that matter to patients?
• How would each possible result change treatment, monitoring, or referral, and is there evidence that those changes help?
• Could false results or unnecessary follow-up lead to harm that outweighs the benefits of using this test or model?
• Were improvements in patient outcomes observed when the test or model guided care, or projected using assumptions in a model?
• Does the evidence compare using this test or model with the care options realistically available to me?